Religious Convictions and Political Choice: Some Further Thoughts

Transcription

1 DePaul Law Review Volume 39 Issue 4 Summer 1990: Symposium - Politics, Religion, and the Relationship between Church and State Article 4 Religious Convictions and Political Choice: Some Further Thoughts Kent Greenawalt Follow this and additional works at: Recommended Citation Kent Greenawalt, Religious Convictions and Political Choice: Some Further Thoughts, 39 DePaul L. Rev (1990) Available at: This Article is brought to you for free and open access by the College of Law at Via Sapientiae. It has been accepted for inclusion in DePaul Law Review by an authorized editor of Via Sapientiae. For more information, please contact wsulliv6@depaul.edu, c.mcclure@depaul.edu.

2 RELIGIOUS CONVICTIONS AND POLITICAL CHOICE: SOME FURTHER THOUGHTS* Kent Greenawalt** INTRODUCTION "Church and state" does not present one unitary question. Instead, there are many linked questions. How close should relations be between government and religious organizations? May government sponsor religious positions and practices? May government aid activities of religious organizations that promote the common good? When, if ever, should religious expression and worship be restricted? When should those with a deeply felt religious objection be exempt from ordinary regulations? Should religious understandings be a self-conscious basis for political choices and dialogue? I have mentioned six questions. The first five are staples of constitutional adjudication in this country under the religion clauses of the first amendment and state constitutions. A rich debate exists about the Supreme Court's approaches to these questions and about the import of our form of government and traditions. The sixth question, however, has received much less attention. Indeed, it has not often been seen as distinct, and legal cases tend not to present the question separate from the others.' While most secular political philosophers have been uninterested in the status accorded to religious convictions in political discourse, those working in religion ethics have assumed that such beliefs properly bear on political choices. Disturbed by modern secular theories of justice that seemed to call for resolution of political issues wholly without regard to transcendent perspectives, I began to think seriously about this question five years ago. I was surprised by how little attention it had received. But today, the picture has changed considerably. Some recent writing develops alternatives with sophistication and persuasiveness, and among that writing are some illuminating * This Article is based on the Seventh Annual Lecture of the Center for Church/State Studies, delivered in Chicago, Illinois on February 15, I am very grateful to members of the Center and the DePaul College of Law faculty for their gracious reception during my visit there. My understanding benefitted from discussion after the lecture, from a presentation to the faculty of New York Law School, and from comments by Robert Audi and Michael Perry. ** Cardozo Professor of Jurisprudence, Columbia University School of Law; J.D., Columbia University School of Law. 1. A statute that requires "creationism" to be taught in public schools if evolution is taught raises the issue of whether religious convictions underlie the legislation. The primary objection to the legislation, however, is that it involves the state in sponsoring a religious viewpoint in public schools. See Edwards v. Aguillard, 482 U.S. 578, 591 (1987). This provides an example of the sixth question arising as an adjunct of the issue of sponsorship. 1019

3 1020 DEPA UL LA W REVIEW [Vol. 39:1019 reviews of my own book on the subject. 2 This occasion gives me an opportunity to clarify my views and place them in the context of recently suggested alternatives. Let me start by putting my topic in a concrete context. Suppose a statute is offered to relieve animals of the oppressively cramped conditions of modern factory farming. Advocates claim that calves, lambs, pigs, and chickens should have a better quality of life before being slaughtered for food. Opponents argue that factory farming helps provide tasty, inexpensive meat and that farmers should be free to decide how to treat animals that they own. At stake in the decision whether to restrict farmers is some balancing of animal interests against human interests. In our relatively wealthy society the human interests are not overwhelming; no one will starve if factory farming is curbed. Among the animal interests, free movement and contact with other animals rank as important considerations. If human interests rightly count for much, much more than animal interests, sacrificing important animal interests for moderate human interests makes sense. In resolving the issue, one must face the question how heavily animal interests should count. Very roughly, people may arrive at conclusions about this question based on reasoned argument and personal feelings. People may also rely on religious convictions they hold, convictions that help set the place of animals and human beings in a broader context. Religious perspectives might lead someone to believe animals deserve great consideration. Conversely, they might lead someone else to suppose that animals exist only to serve human welfare. We can quickly see how different religious perspectives could affect views about restricting factory farming. Do these religious convictions have an appropriate place in our political life? Before responding, I need to draw two distinctions. One is between private individuals and public officials. Perhaps civility in a pluralist liberal democracy only concerns public officials and their actions. The other distinction is between mere reliance on religious convictions and public argument in those terms. Perhaps religious arguments are inappropriate, even if votes or political activities based on religious grounds are valid. A comprehensive approach to the larger question, namely whether religious convictions should play a part in our political life, must address citizens and officials, reliance and public argument. I. RELIANCE ON RELIGIOUS CONVICTIONS Before I present my own position, I want to say a little more about what relying on religious convictions means. No one doubts that political judgments and arguments can be influenced by religious convictions. In fact, people are influenced in what they do and say by all of the important aspects of their lives. But people should make an effort to discount some influences, 2. K. GREENAWALT, RELIGIOUS CONVICTIONS AND POLITICAL CHOICE (1988).

4 1990] SOME FURTHER THOUGHTS 1021 such as prejudices that they recognize as ill founded. Should they also try to discount religious perspectives in favor of bases of judgment that everyone in society can share? That is the question.' Some of you may be wondering how well suited this topic is for an Annual Lecture of the Center for Church/State Studies, sponsored by a part of a university thought of as Roman Catholic. 4 Roman Catholicism has a rich tradition of natural law that emphasizes what common human reasoning can discover about morality and political justice. Natural law arguments appeal to all people capable of reasoning; they do not depend on particular religious commitments. Any worry about whether people should rely on religious perspectives seems to have no application to these arguments. But matters are not so simple. Many Roman Catholics accept the authority of the Church on social issues, even though they are not fully persuaded by the naturalist arguments that correspond with the Church's position. For some of those persuaded by naturalist arguments, confidence that they are correct is increased by authoritative declarations. Further, one's degree of certainty can affect what one is willing to impose by legal regulation.' There is, however, another, more subtle point. Religious belief and naturalist reasoning are sometimes interwoven in a way that is not usually recognized. 6 Consider, for example, the status of the fetus. Does the moral consideration it is owed grow steadily over time, or is there a single point when it first deserves consideration and also deserves full consideration? The idea that God ensouls human beings may lead one to look for a single point of critical importance. Naturalist reasoning may determine when that event occurs, but religious underpinnings may incline one away from the idea of steady growth in moral status. For Roman Catholics, the question about reliance on religious convictions is less stark than for those Protestants and Jews who think that some moral and political truths, not discoverable by reason, can be read directly out of biblical passages. Nonetheless, the question matters for Catholics, as well as Protestants and Jews. A. The Place of Religious Convictions In Political Choice and Argument The extreme answers to the question concerning reliance on religious convictions are straightforward. One is that citizens may properly rely on 3. No problem is raised if religious belief has led someone to a position he now thinks he can fully and persuasively defend in nonreligious terms. The issue is not about the causal relation between religious belief at some earlier time and one's present position; instead, the issue concerns one's present dependence on religious conviction to support his position. 4. The DePaul University College of Law sponsors the Center for Church/State Studies. DePaul University is affiliated with the Vincentians, a Roman Catholic order. The Center itself is strongly nondenominational. 5. I discuss these matters in greater depth in K. GREENAWALT, supra note 2, at See id. at

5 1022 DEPA UL LA W REVIEW [Vol. 39:1019 and argue in terms of any sources of truth they believe are valid, including their religious convictions. The other is that political life should be carried on in terms of reasons that all citizens can appreciate, and this excludes religious convictions. Inquiring about the responsibilities of a good citizen of our liberal democracy, I take an intermediate position concerning reliance upon religious convictions. I believe that good citizens should not seek to forbid behavior just because they think it sinful or intrinsically wrong. The only basis for a legal restriction is a harm or good that is comprehensible in secular terms. I shall pass over this point quickly, because it does not help resolve most problems, including our factory farming example. The harms to animals that restrictions on factory farming would prevent are perfectly comprehensible in secular terms. With respect to such issues, the good citizen should remain open to reasons that all members of society can evaluate. If those reasons yield a decisive answer, the citizen should accept it. Nevertheless, shared principles of justice, shared methods of assessing values, and shared ways of determining facts will often prove inconclusive. The problem is not solely that considerations on each side may be neatly balanced. Rather, common reasoning may be radically inconclusive, as seems to me true about the status of animals. Although we have some idea about how "higher" animals resemble human beings, we disagree sharply about how much human beings owe animals. Moreover, the right answer does not seem susceptible to reasoned analysis by itself. In reaching a conclusion, everyone must rely on a sense of the place that nonhuman creatures inhabit in our world that is not fully based on shared premises or ways of reasoning. If this much of my account is accurate, it follows that people cannot be expected to rely exclusively on shared premises and ways of reasoning. If people can appropriately rely on personal intuitions and perspectives, they should also be able to rely on religious perspectives. Most religious persons could hardly decide what personal perspectives they would have but for their religious beliefs. And there is no sufficient reason to privilege nonreligious personal perspectives in preference to religious ones. If, as I believe, good citizens can properly rely on religious convictions for many political issues, then public officials can reasonably rely on judgments formed in this way by their constituents or other citizens. Further, officials will often be in the same position as citizens. Sometimes when they exercise their own judgments they will have to rely on sources of understanding that are not commonly shared; and religious sources should be regarded like other such sources. My view about political discourse, however, is different. The actual debate of political issues in terms of competing religious convictions is disturbing in a pluralist society. Civility and respect for minorities counsel that public advocacy be conducted in the nonreligious language of shared premises and modes of reasoning. Such a statement most obviously holds true for argument by public officials and candidates, but I think it is also true for public advocacy by private citizens. For example, a column or letter in the New

6 1990] SOME FURTHER THOUGHTS 1023 York Times should not contend that restrictions on factory farming are warranted because they conform with particular biblical passages. The combination of my views leaves a disturbing discrepancy between the reasons people may have and rely upon, and the reasons they should assert in public advocacy. Because public advocacy is not generally thought to involve full disclosure and analysis of all one's reasons for a position, however, this discrepancy does not involve deception and hypocrisy; or so I claim. This is the bare bones of my position. It is punctuated by doubts and qualifications, and some of these will be evident in what follows. But what I have said provides enough context to consider opposing views that have been recently advanced. B. Relevance for Religious Persons: Conceptions of Liberalism Before I proceed to consider two major alternatives to my view, I want to explain the model of good liberal citizenship that I have in mind and why this model might matter to a person whose religious beliefs provide sources of understanding for ethical issues. Some recent discussions in political philosophy have drawn a distinction between liberalism as a comprehensive conception of human morality and liberalism as a political conception.' Liberalism as a political conception does not address many controversial questions about what is good or how moral truth in general is to be discovered. Rather, it focuses on political institutions in a pluralist society where many views on these subjects are held. Persons who believe that the best lives are lived voluntarily in closely knit, authoritarian religious communities are not liberals in general, but they may be political liberals. That is, they may believe that the political institutions of liberal democracy are best suited for the flourishing of this best life and other good lives. The distinction between a comprehensive liberal view and a political conception of liberalism has been elaborated by John Rawls in response to criticisms that his work, A Theory of Justice, 8 presupposed an impoverished liberal view of human beings. Rawls has answered that he intends his theory only as a political conception. 9 My observations about reliance on religious convictions also concern a political conception of liberal democracy. I do not propose any general theory about human good or moral truth. Indeed, since I concentrate on only one aspect of liberal democratic theory and rely on our own country's traditions and modern social environment, my claims are much less general than those found in a full liberal democratic theory such as Rawls's. While it is true that I offer guidance to citizens as they make up their minds about 7. See C. LARMoIR, PATTERNS OF MORAL. COMPLEXITY 1-90 (1987); Rawls, The Idea of an Overlapping Consensus, 7 OXFORD J. LEGAL STuD. 1, 2-15 (1987). 8. J. RAWLs, A THEORY OF JusTIcE (1971). 9. See J. RAwLs, supra note 8, at 2-15.

7 1024 DEPA UL LA W REVIEW [Vol. 39:1019 public policy, that does not mean, as one commentator has concluded, 0 that I am reaching beyond a political conception to a "comprehensive" liberal conception. Let me explain why. We can imagine a liberal political theory that limits itself to institutional structures. Under such a theory, given proper institutions, citizens and officials could decide and argue on any bases they chose. But modern theories of justice are richer than this. They assert regulative principles of justice for liberal democratic societies that reach purposes and grounds as well as outcomes.ii To take an obvious instance, a political aim to subordinate members of a minority racial group would be unjust because such an aim contravenes the premise of equal citizenship that lies at the heart of liberal democracy. A politician who openly argued that all preferential treatment for blacks should be stopped merely because continued subordination of blacks is desirable would violate liberal principles of justice. If such an argument should not be made by a politician, neither should it be made by a liberal citizen speaking or writing to a public audience. With respect to officials, it is easy to see that liberal principles do not stop at advocacy. Even if they were silent about their motives, officials who acted to end all preferential treatment because they wanted subordination of blacks to continue would 10. Tushnet, Religion in Politics (Book Review), 89 COLUM. L. REV. 1131, 1136 (1989). Professor Tushnet says, "Greenawalt, however, treats liberalism as a general and comprehensive conception, in the sense that liberalism, as he sees it, offers guidance to individual citizens as they make up their minds about controverted questions of public policy." Id. 11. In treating my own view as comprehensive, Tushnet differentiates it from that of Rawls. Id. Tushnet writes "[a]ccording to Rawls's version of liberal theory there are no criteria that limit the bases of a good liberal citizen's political action; that is, in his version, liberalism is a political theory and not a general theory of good behavior." Id. (emphasis added). See also id. at Tushnet is correct that Rawls does not present a general theory of good behavior. I do not either. But Rawls's theory clearly does constrain how citizens are to act with respect to many political subjects. Rawls contends: Justification in matters of political justice is addressed to others who disagree with us, and therefore it proceeds from some consensus: from premises that we and others recognize as true, or as reasonable for the purpose of reaching a working agreement on the fundamentals of political justice.... [Clonnected with a political conception of justice is an essential companion conception of free public reason. This conception involves various elements. A crucial one is this; just as a political conception of justice needs certain principles of justice for the basic structure to specify its content, it also needs certain guidelines of inquiry and publicly recognized rules of assessing evidence to govern its application... [G]iven the fact of pluralism, there is, I think, no better practical alternative than to limit ourselves to the shared methods of, and the public knowledge available to, common sense, and the procedures and conclusions of science when these are not controversial. Rawls, supra note 7, at 6-8.

8 1990] SOME FURTHER THOUGHTS 1025 violate principles of justice. 2 Once it is understood why most political conceptions bear on how officials make decisions, their relevance for how ordinary citizens make political decisions follows closely. People vote in referendums, and they also elect officials. Officials often defer to constituency opinion. Good citizens should try not to reach decisions about political issues on grounds that violate liberal democratic principles. It is in this spirit that I explore not only public arguments, but also the manner in which officials and citizens try to reach judgments. More might be said about this, 3 but I hope it is clear why my attention to choice, as well as argument, does not mean that I am moving beyond a political conception of liberal democracy to a comprehensive liberal doctrine. A more troubling problem involves the practical point of worrying whether a model of good liberal citizenship might preclude some religiously based grounds for political decisions. For the person with religious beliefs, those beliefs are paramount; if they conflict with some model of liberal democracy, too bad for the model. A person who knows what his religious beliefs indicate about a political issue is going to follow those beliefs in any event. So, asking what good liberal citizenship implies about reliance on religious beliefs may make little sense. What audience of religious believers is there to be influenced by such an inquiry? 4 Even if there were no audience to be influenced, there would be some value in discussing where the premises of liberal democracy and certain religious perspectives may conflict. Further, one should not oversimplify the relation between religious convictions and political principles. Not everyone who has religious beliefs accords them the paramountcy in their lives that the beliefs call for. A person persuaded that his religious views actually conflicted with ideals of liberal democracy might disregard those religious views. More subtly, a person persuaded of the merits of liberal democracy might use ideals of liberal democracy as a test to determine which religious convictions to accept.'" He might reject religious perspectives that seemed sharply at odds with political ideals he accepts, or given acceptance of some broad religious perspective, he might interpret that perspective in a way to minimize possible conflicts. Each of these points has some merit, but none 12. Indeed, they would also violate the constitutional principle of equal protection, although perhaps not in a manner that courts could discern. 13. Tushnet points out that I say liberalism includes acceptance of a modest degree of individualism and rationalism. Tushnet, supra note 10, at See K. GREENAWALT, supra note 2, at That acceptance might be understood as being limited to political subjects or as indicating that some extremely "thin" components of a more comprehensive view about human capacity and moral choice are needed for the belief that liberal democracy is a highly desirable form of government. 14. The way I have put this challenge is starker than in any criticism I have seen, but something like it is found in Gamwell, May Citizens Be Religious?, (Book Review), J. L. & RELIGION (forthcoming). See also Tushnet, supra note 10, at 1131 (questioning the relevance of my book). 15. See Rawls, supra note 7, at 19.

9 1026 DEPA UL LA W REVIEW [Vol. 39:1019 reaches the heart of the matter, which concerns the levels of content in many religious perspectives. Some religious convictions do lead to political positions that are at odds with my claims about liberal democracy. For example, if someone thinks biblical passages indicate God's abhorrence of homosexual acts and that God wants all societies to forbid them, that person will support restrictive legislation, whatever the implications of liberal political premises. But many religious perspectives accommodate distinctions between behavior that is "wrong" and behavior that should be restricted by law. Further, many religious perspectives are positively committed to religious liberty and to separation of church and state. Thus, people with such beliefs may well conclude that they should not impose their religiously based views of what is wrong on others who do not accept their religion. They may think that such wrongful behavior should not be forbidden to citizens who reasonably cannot be persuaded of their wrongdoing. Other religions suppose that an ideal society would be dominated by its own perspectives but that in a second-best religiously pluralist society, religious liberty and nonestablishment are appropriate conditions of participation, making comfortable coexistence possible. Adherents of such religions might also decline to act politically on some moral truths religiously established. For people whose religious perspectives include some accommodation to religious liberty and nonestablishment, the implications of those political concepts can matter. They can affect how far believers try to enforce religiously grounded ideas of proper behavior. That is the audience of religious believers I am trying to reach, and I believe it to be rather large. II. RELIGIOUS BELIEFS As A SUBJECT OF POLITICAL DEBATE Some scholars who have agreed with me that people often appropriately rely on religious convictions in making political decisions go further and urge that such convictions should commonly be a part of political dialogue. 16 Their challenge to my contrary position is composed of at least three threads. The first is that religious and other bases for political judgment are so intertwined that attempting to excise religious bases from political argument does not make sense. The second is that religious grounds may not be distinguishable from other grounds in the degree to which they are "publicly accessible" or rest on rational support. The third thread is that political and cultural life will benefit from a full airing of religious, as well as other, grounds for political positions. In considering this challenge to my approach, it may help to consider an example drawn from my book, where I distinguished the acceptable use of 16. See M. PERRY, MorLITY, POLITICS, AND LAW (1988); Gamwell, supra note 14; Gamwell, Religion and Reason in American Politics, 2 J. L. & RELIGION 325 (1984) [hereinafter Gamwell, Religion and Reason); Lovin, Perry, Naturalism, and Religion in Public, 63 TuL. L. REv (1989); Perry, Comment on "The Limits of Rationality and the Place of Religious Conviction: Protecting Animals and the Environment," 27 WM. & MARY L. Rav (1986).

10 1990] SOME FURTHER THOUGHTS 1027 religious imagery from an explicit reliance on religious grounds, the latter of which I claim is inappropriate. 17 Two speakers, opposing cuts in welfare programs, use the story of Cain and Abel to support their positions. The first speaker says: Am I my brother's keeper? Time and again that question presents itself to us as individuals and a society. The answer is yes. We cannot be true to ourselves and our traditions if we fail to cope adequately for the poor and needy. 8 The second speaker says: The story of Cain and Abel as reported in Genesis clearly establishes that God wants us to take care of other people; other parts of the Old Testament (citing passages) and the continuous teachings of the true Christian church (citing authoritative church documents) reveal the minimum amount of care for the poor and needy that is required; the proposed cutbacks fall under those minimum standards and are therefore contrary to the will of God. 9 How far apart from me are those who recommend a greater infusion of religious perspectives in political dialogue? Without comment on more illustrations it is hard to tell; but the three writers I shall mention believe they disagree with me, and would apparently accept much more political argument cast on religious premises than would I. Michael Perry, a leading constitutional scholar who is also a Roman Catholic, has criticized my suggestion that public political dialogue should largely be free of explicit religious claims. 20 He urges "a deliberative, transformative politics... in which questions of human good, of what way or ways of life human good consists in, are not marginalized or privatized but, instead, have a central, public place." ' 21 Further, he claims that: [Olne can participate in politics and law... only as a partisan of particular moral/religious convictions... [1If politics is and must be in part about the credibility of such convictions, then we who want to participate, whether as theorists or activists or both, must examine our own convictions self-critically. We must be willing to let our convictions be tested in ecumenical dialogue with others who do not share them. 2 Recommending Perry's approach in preference. to my own, Robin Lovin criticizes me for taking political commitments as set, instead of open to evaluation and transformation. 23 He emphasizes the complexity of relating religious convictions to political choices. Lovin asks whether an opponent 17. K. GREENAWALT, supra note 2, at Id. 19. Id. 20. Perry, supra note 16, at M. PERRY, supra note 16, at Id. at 183. Perry's views are more fully developed in a work-in-progress, tentatively titled "Love and Power: A Postliberal Reflection on Religion, Politics, and Human Rights." 23. Lovin, supra note 16, at

11 1028 DEPA UL LA W RE VIEW [Vol. 39:1019 of nuclear deterrence should say: "My faith tells me that we cannot continue to rely on nuclear weapons for security without betraying a more fundamental reliance on God."1 2 4 He suggests 25 that believers use religious language in political discussions to proclaim an alternative way of ordering human life, to convert secular citizens to their religious premises, and "to articulat(e] an idea of the human good ' 26 and how it might be achieved. Like Perry, Lovin identifies himself with the tradition of ethical naturalism and argues that one benefit of public discourse about religious premises is that it "opens the way... for the recasting of religious beliefs in light of other, nonreligious knowledge.' '27 Franklin Gamwell concentrates his challenge on the nature of religious belief. 2 " Notwithstanding the prevalent appeal of the idea that religious convictions are inherently nonrational or beyond reason, Gamwell rejects this proposition, finding "no persuasive case for this understanding of religion." ' 29 Like Franklin, Jefferson, and Madison, Gamwell thinks religious convictions are subject to rational examination and are not beyond public scrutiny. Only convictions that are subject to rational review have a claim on citizens, but this category includes many religious convictions. Arguing that public debate should be as full as possible, Gamwell states that "[t]o seek the political implications of one's faith, and to advocate both in the public debate, is not only permitted to but incumbent upon citizens of the Republic." 30 The idea that religious convictions themselves may be a proper subject of political debate is appealing. Not only does it respect the complexity with which religious convictions intertwine with political judgments, it also avoids any disturbing discrepancy between the reasons people properly have and 24. Id. at Id. at Id. at Id. at See Gamwell, supra note 14, at 13. He also contends that my position is not only mistaken but skirts logical error in relying on publicly accessible grounds to show the limits of rational grounds. I agree with Dean Gamwell that rational thought cannot show the validity of "nonrational" sources of understanding. If rational thought could show the validity of a source of understanding, the source would be rational. I also agree that one cannot establish beyond doubt that rational arguments, or publicly accessible arguments, are inconclusive on some major political issues. I claim that when we examine those arguments in the context of issues like animal rights and abortion, they seem radically inconclusive. Further, I contend that if publicly accessible, or rational, arguments are now radically inconclusive, as they seem to be, citizens will have to rely on some other bases of judgment. Although I believe on faith that what I shall loosely call "nonrational" sources can yield valid insights, my belief is not a requisite of my claims about liberal democracy. If citizens must inevitably rely on nonrational sources, their reliance is appropriate or valid, whether or not the sources of judgment yield true insights. 29. Gamwell, Religion and Reason, supra note 16, at See Gamwell, supra note 14, at Gamwell, Religion and Reason, supra note 16, at 339.

12 1990] SOME FURTHER THOUGHTS 1029 the reasons they properly advance in political argument. My book was directed primarily against the idea that people should try to decide political issues without reference to religious convictions. 3 I may have given less attention than was warranted to the claim that those convictions should also enjoy a full place in political debate. The challenge from that direction makes me realize that my own views need clarification and further development. My suggestion that religious convictions should, for the most part, not be explicitly argued in political debate is not based on some overarching first principle of liberal democracy. If I grant these convictions an appropriate role in making many decisions, as I do, there must be special reasons to limit their place in political discussion. Much depends on the nature of religious convictions, and the place of religious beliefs and religious identification, in our society. Gamwell explicitly claims, and the sense of Perry's and Lovin's discussion is similar, that religious convictions are not more or less the subject of rational evaluation than other bases of political judgment. 32 If that were true, it would constitute a substantial argument in favor of religious convictions being an explicit aspect of political debate. In assessing Gamwell's claim that religious convictions are subject to rational evaluation, I shall concentrate on Christian convictions. They are strongly dominant in our culture. I am reasonably confident that what I say about these applies to Jewish convictions and to many other religious convictions. A. Bases of Religious Convictions One can roughly distinguish five bases for religious convictions. 33 The first two bases can be explained in a fairly straightforward manner. The fourth and fifth bases, both derivative in nature, build upon the other bases. I will discuss the third basis, namely personal experiences, in greater detail because it is central to my contention that religious beliefs are not accessible to public reason. The first is a basis that is supported directly by thought about human beings in their social existence. Direct naturalist beliefs, which we might call moral-religious, rest on such bases. Many beliefs about human good and right human action are thought to flow from a reasonable understanding of what it is to be human. These can be evaluated by someone of radically different religious belief. The peculiar religious aspect of the belief may provide an extra degree of confidence, may color the content of the belief, 31. K. GREENAWALT, supra note 2, at 4, See Gamwell, Religion and Reason, supra note 16, at ; Gamwell, supra note 14, at 16-18; Lovin, supra note 16, at ; Perry, supra note 16, at I recognize that other bases may exist or that one might classify these bases differently; but this loose identification of bases helps clarify the issues discussed in this Article.

13 1030 DEPA UL LA W REVIEW [Vol. 39: and may give the belief an extra weight or magnitude' but the natural reasons behind the belief may be understood and assessed by all reasonable persons. It is no coincidence that Perry and Lovin are adherents of a naturalist ethics and politics. That perspective, as Lovin says," 5 fits most comfortably with an absence of restraint on religious claims in the political process. A second basis for religious conviction is one that is claimed to be supported directly by a rational understanding that reaches beyond human good and human society. Philosophical arguments for the existence of God and about God's qualities are examples. They offer themselves for acceptance or rejection on reasoned bases that are said to be widely accessible. A third basis for religious belief is what I shall call tradition or personal experience. I am not talking about tradition or personal experience as inevitable influences on what people believe; I am talking about them as self-conscious grounds for belief. A person might say, "I have correct beliefs because I was fortunate to be born into a particular tradition, but I have no arguments that might persuade persons born into another tradition that the perspectives of my tradition are correct." I shall pass over the problem of the possible circularity of a belief put so starkly and pay attention to the more interesting problem of personal experience. A common basis for Christian belief is what I shall call confirmation by personal experience. The inadequacy of other broad perspectives and the fulfilling quality of Christian life are supported by one's personal experiences: one feels in the depths of one's being that the message of a loving God revealed in the life of Jesus is true.1 6 I think the periods of doubt that accompany belief are closely related to the importance of this kind of confirmation. My brother-in-law recently preached a sermon on what he called existential atheism. He noted that in the aftermath of devastating personal experiences, most people find it difficult to believe that a loving and just God exists.1 7 Afterward, members of his congregation overwhelmed 34. For example, it may be thought that naturalist reasoning can show that it is always wrong to act intentionally to take the life of an innocent person. Belief in a providential God may give one confidence that this principle is sound even as applied to many cases when more lives would be saved if an innocent person were intentionally killed. Belief in God may also lead one to think that taking an innocent life is a particularly grave wrong because it usurps God's responsibility for life and death. See Greenawalt, Natural Law and Political Choice: The General Justification Defense; Criteria for Political Action; and the Duty to Obey the Law, 36 CATH. U.L. REV. 1, 6-23 (1986) (the absolute principle is hard to defend on naturalist grounds alone). 35. Lovin, supra note 16, at discuss feelings and beliefs in a highly simplified manner here. For any one individual, obviously much more would need to be said about what he or she believes and what feelings support that belief. 37. Sermon by William Abernethy, Existential Atheism: The Personal Challenge, The Wellesley Congregational Church (1989). See also C.S. Lawis, A GRaF OBSERVED (1961) (on Lewis's doubts and sense of meaningless after the death of his wife).

14 1990] SOME FURTHER THOUGHTS 1031 him by sharing their own feelings of doubt. Terrible circumstances in life can make everything ordinary seem trivial, but they leave unshaken many of our beliefs, including our ordinary ethical beliefs. 5 In trying circumstances, we do not begin to doubt principles of physics, conclusions about historical events, or the human value of love. Does it follow that the sorts of beliefs that are susceptible to grave doubt when life is unbearably hard are, in comparison, illusions? That is not the conclusion I would draw. I would say rather that religious beliefs involve one's whole being and experience in a special way, and that susceptibility to existential doubt is an aspect of much religious belief. Of course, what I say about this does not matter much, but it does matter that this perspective corresponds with much of the Christian tradition. Here we reach a critical question about Dean Gamwell's understanding of religious truth. He speaks, for example, of "laws which are informed by the truth that is accessible to common experience and reason." 3 9 What is the status for him of an individual's personal experience about religion? First we need ask whether that experience is explicable to someone else. To me, it seems largely explicable. Communicating aspects of it to people who have never had such experiences may be like trying to tell someone who has never been in love what that feeling is like, really beyond the reach of understanding of the other person. But a sense of shallowness and meaningless of life is common, as is a quiet desperation about one's own personal life. Even those with no religious experience will have had moments of fulfillment and joy when things seem to "come together," so someone's saying that religious belief and practice confer fulfillment and contentment will not be incomprehensible to others. Even if a person's religious experience is commonly accessible, in the sense that it is explicable, a big problem remains. The person who has actual experiences that might be relied upon will give them far more weight than will those individuals that he tells about the experiences. That is natural. When a believer offers personal confirmation as evidence of the truth of his religious beliefs, that confirmation serves as some evidence for others, but it is much weaker evidence than it is for the believer. There is, then, no interpersonal way in which the weight of personal experience is to be assessed. If a law were based largely on religious beliefs that were mainly confirmed by personal experiences, those who had not shared in the experience might understand why the law had been adopted, but to them, there would be no reasoned basis on which they would be able to conclude that the law was sound. Someone like Dean Gamwell, who claims that religious claims are accessible to common experience and reason, might respond to this problem in 38. That, at least, is my own personal experience after the death of my beloved wife, Sanja; and it is my sense of the experience of others I know. 39. Gamwell, Religion and Reason, supra note 16, at 338.

15 1032 DEPA UL LA W REVIEW [Vol. 39:1019 various ways. First, he might say that as citizens dealing with political subjects, believers should give no more weight to their own personal experiences than would a reasonable outsider. 4 0 That, however, would require mental gymnastics of an -extreme sort. Instead, he might think that such personal confirmation by experience simply proves to be an inappropriate ground of religious belief. If so, he would stand in opposition to much of the Christian tradition. Finally, he might claim that explicability is sufficient to assure that a conviction is based on rational reflection. But if explicability alone were enough, we could not expect that "full public debate" would provide a very helpful test of what beliefs should be held. 41 This complicated reliance on personal experience, and tradition, that I have described is largely what I mean to capsulize by saying that religious beliefs are largely "nonrational" or "not accessible to public reason." I do not think religious convictions as a whole can be lumped into a big category called nonrational. Reasoned judgment plays an important part in what most people believe about matters of religion. Nevertheless, I do think that people's religious convictions rest partly on elements that are not subject to reasoned interpersonal evaluation. For purposes of this discussion, whether confirmation by personal experience is a source of true religious understanding is not critical. What is critical, rather, is that many members of our society regard it as such. What may also be important is that some people, who naively think that a more solid grounding in reason exists for their beliefs, would learn upon closer examination that the reasoned arguments are less decisive than they have supposed. I turn now to the fourth and fifth bases for religious understanding. These are bases that I loosely call derivative 42 because they rest wholly or largely on other bases. For example, someone who believes that every passage of the Bible is the infallible word of God must think that the existence of God and the authority of the Bible are somehow established. Someone might think the underlying bases can all be rationally established. Perhaps the existence of God can be shown by philosophic argument: the authority of the Bible shown by miracles and the accuracy of prophecies. Or, a person may think that confirmation by personal experience, or tradition, is needed. In my categorization of bases, the fourth basis for religious understanding is one that is not directly rationally establishable, but is thought to be rationally establishable in light of other bases that are rationally establishable. The fifth basis is one that is derivative from other convictions, some of which are understood not to be rationally establishable. 40. Someone who took this position might go further still and say that for all purposes of assessing truth, a person should similarly discount his or her own personal experience. 41. Id. at It is possible that bases might link together rather than some being derived from others. I will omit the extra complexities introduced by this possibility, but I do not think they affect the analysis in any crucial way.

16 1990] SOME FURTHER THOUGHTS 1033 B. Application of Religious Convictions to Political Dialogue Of all these kinds of religious convictions, only direct naturalist claims seem likely to be a fruitful subject of exchange in political debates over legislation and public policy. I have suggested that interpersonal reason is sharply limited when one relies heavily on personal experience or tradition. That difficulty obviously extends to any religious convictions that are derived indirectly from these sources. For example, if a political advocate says, "this passage of the Bible establishes that the interests of animals do not count, and I know by faith that the Bible is infallible," there is not much opportunity for genuine dialogue with someone who lacks that faith. Suppose instead the advocate thinks biblical infallibility can be established by a series of rational steps, including philosophical arguments about God's existence, miracles, and the accuracy of prophecy. Now the speaker's claims are subject to judgment by interpersonal reason; but a political debate over the treatment of animals is hardly the occasion to run through each of these claims step by step. Casting doubt on the reasoned basis for biblical infallibility will not appear productive to those whose aim is to protect animals. They might feel that even if they succeeded, the speaker would probably cling to infallibility, giving up only his belief that it could be fully supported on rational bases. What genuine dialogue occurs when particular biblical passages are relied on as support for political positions is mostly between people who interpret the passages differently. The problem of useful interaction extends to philosophic arguments about God's existence. A debate that primarily concerns political policy is hardly the best forum to examine those arguments. A more likely result is the trading of insults: "what can you expect from someone who rejects even the obvious existence of God?" and "anyone foolish enough to think God's existence can be shown lacks judgment." The point of connection between those who think very different things about religious propositions is greatest when the focus is on what our reason tells us about human good and ethical action. Here is a genuine opportunity for mutual understanding and reciprocal persuasion. How attractive are the recommendations of Perry, Lovin, and Gamwell in light of these different sources of religious convictions that affect political judgment? If Gamwell wants to admit into public debate all religious convictions that are subject in any degree to reasoned analysis, then even those based substantially on confirmation by personal experience would be appropriately argued. But such an outcome would clutter political debate with many religious claims that are not subject to reasoned interpersonal evaluation. Suppose, instead, Gamwell would encourage reliance only on claims believed to be establishable by reasoned interpersonal argument. This would still include first, claims about God that rest on philosophic arguments and second, derivative claims like the infallibility of biblical passages or the authority of a religious leader, when those claims rest upon links in a chain of reasoned argument. As I have said, political debate is not a good vehicle for addressing such claims.

17 1034 DEPA UL LA W REVIEW (Vol. 39:1019 Further, there might well be a disturbing discrepancy in what religious believers could argue based on their philosophical sophistication. Imagine two people with similar religious views. The first might think that all his beliefs could be grounded in interpersonal reason; the second, acquainted with competing arguments, might understand the crucial limits of the rational arguments and the crucial place of confirmation by experience. The first, but not the second, would appropriately make his religious arguments in the political domain. The position of Perry and Lovin apparently is that all religious underpinnings of political positions would properly be a part of political debate. But that conclusion has some paradoxical aspects. In his book, Perry suggests that "constitutional adjudication is, at its best, a model of deliberative, transformative politics. ' 43 We do not expect judges in constitutional cases to indicate direct reliance on religious premises. Perry's discussion of constitutional adjudication does not suggest that this is a defect." But if constitutional adjudication is a model of transformative politics and it can appropriately be carried on without explicit reliance on religious premises, why may that not also be so about political arguments concerning legislation and public policy? If fruitful dialogue will occur when naturalist premises are relied upon, and those premises can be understood and evaluated by persons of different religious beliefs, how much will be lost if the religious convictions accompanying the naturalist premises are not fully developed? Robin Lovin ends his essay with the interesting observation that religious believers who reject ethical naturalism will not agree with his understanding of religion and public discourse. 4 5 But his recommendations for political dialogue apparently reach all religious convictions. If one thinks in those terms, one must ask about the effects of all the kinds of religious claims that are likely to be advocated. Although my book talks mostly of the presentation of arguments in the political process, I did not mean to exclude the possible transformation of views in dialogue. I believe that both political judgments and religious understandings can be enriched and altered by interchanges with people of different views. I am, however, less optimistic about the benefits of public dialogue than Perry, Lovin, and Gamwell. I am particularly skeptical about the promise of religious perspectives being transformed in what is primarily political debate. That skepticism alone would not justify my suggested exclusion of explicit religious arguments from ordinary political advocacy. Much depends on present tolerance and whether certain views about religion are prevailing. If virtually everyone shares a religious belief, then argument in terms of that 43. M. PERRY, supra note 16, at Id. at See also Lovin, supra note 16, at Lovin speaks of the law as having "its own highly circumscribed definition of 'publicly accessible' reasons," and obviously believes it would be undesirable if general political dialogue were not much richer. Id. 45. Lovin, supra note 16, at 1538.

18 1990] SOME FURTHER THOUGHTS 1035 belief may be untroubling. Ironically, if very few people have the belief, argument in terms of it may also be untroubling, because no one will suppose policy is actually made on that basis. What is troubling is argument in terms of a belief that many share but many others do not. Imagine, for example, a state within this country where many Protestants believe in biblical infallibility. If political argument were comprised largely of debate about the meaning of particular biblical passages, those who did not believe in the same kind of authoritativeness for biblical passages would be bound to feel excluded. My own recommendations about political discussion are grounded largely in concern about such situations. I find myself unable to accept all that Perry, Lovin, and Gamwell recommend. I remain convinced that some screening of religious convictions is appropriate for political debate. But what I wrote in my book on this subject may be insufficiently nuanced. I do not think religion is just a "private affair." Nor do I think it is just a "community affair" within limited voluntary communities. Talk about religion is important for people to understand and evaluate our culture and the meaning of their own lives. 4 6 Considering the extent to which Americans regard themselves as religious, the marginalization of religion in our culture's intellectual life and in public media is somewhat surprising, and I believe it is regrettable. One aspect of developing religious perspectives before a broader public is to suggest their implications for political issues. But that is different from engaging in direct political advocacy of particular laws and policies. It is for direct advocacy that I have suggested that religious convictions not be presented as arguments in public debate. The dangers of divisiveness and exclusion are greatest when public officials explicitly rely on religious grounds. 47 Their doing so will, of course, have the benefit of giving voters a fuller sense of what motivates their representatives. But the costs of policies being debated in terms of competing religious premises outweigh the benefits at this stage of American history. Moreover, representatives can fairly present their characters and personalities without speaking explicitly in terms of religious premises when they deal with political issues. Former president Jimmy Carter is an apt exemplar in this respect. 46. Id. at To illustrate the religious aim of proclamation, Lovin talks of a Mennonite farmer who declares his religious reasons for supplying fresh vegetables to soup kitchens. Id. at Nothing I say casts doubt on the appropriateness of such explanations. When the religious aim is conversion, id. at , I also agree that explaining one's religious perspectives for political judgments is proper. Finally, the third religious aim Lovin discusses is the articulation of an idea of the human good. Id. at I agree with him that there is nothing wrong with speaking convictions that bear on this in public settings. I question only the development of explicitly religious arguments as an aspect of debates about particular laws and policies. Thus, exactly how great the practical distance is between Lovin and myself is unclear to me. Would he, or Perry or Gamwell, view with equanimity speeches in legislatures developing elaborate theological bases for political positions? I am not sure. 47. K. GREENAWALT, supra note 2, at ,

19 1036 DEPA UL LA W REVIEW [Vol. 39:1019 He made his religious persuasions clear to the public but did not argue for specific policies in explicitly religious terms. Because private citizens are not representatives of the whole public, their arguments based on explicit religious grounds are less disturbing, but I still think that people who are not religious leaders should generally rely on nonreligious arguments for particular laws and policies. This caution is especially important for private citizens with substantial political influence. The less vague the religious premises and the less closely they connect to views that are not dependent on the religious premises, the greater the dangers of divisiveness and exclusion. 4 Although drawing sharp lines may be impossible, the development of explicitly religious premises is more appropriate for discourse in our general culture than for direct political advocacy. III. SECULAR, PUBLICLY ACCESSIBLE BASES FOR DECISION My book was directed mainly against the idea that citizens and officials should try to make all political decisions on the basis of reasons accessible to all, reasons that do not include any religious premises. 49 Variations of that view have recently received powerful defense in articles by Thomas Nagel 5 0 and Robert Audi 5 and in reviews of the book by Audi 2 and David Richards." A. Views That People Should Try To Exclude Religious Convictions From Political Choices Nagel suggests that political liberalism rests on what he calls a higherorder impartiality. 5 4 Treating all people impartially within one's own view is 48. As a conversation with Michael Perry led me to see, an advocate of including religious convictions in political dialogue might take something like the following position: The best modern theology tries to understand a religious tradition in light of all existing sources of understanding. See, e.g., D. TRAcY, BLESSED RAGE FOR ORDER (1988). There is really no sharp break between religious premises and other avenues of insight. The right kind of religious understanding poses minimal dangers of divisiveness and exclusion. I argue only that discussion in those terms is appropriate for political debate. In this Article, and in my book, I do not in my recommendations concerning political argument distinguish among kinds of religious positions. For a principle of political philosophy, it seemed unwise to propose that some religious premises should be argued but not others. I am still inclined to that view; but it is undeniably true that the argument of some kinds of religious convictions may pose greater harms, and yield lesser benefits, than the argument of others. An example concerning literalist biblical assertions is not a fair measure of the effect of religious understandings of the sort suggested above. 49. K. GREENAWALT, supra note 2, at Nagel, Moral Conflict and Political Legitimacy, 16 Pm. & PUB. AsF. 215 (1987). 51. Audi, The Separation of Church and State and the Obligations of Citizenship, 18 PHm. & Pua. Ai'. 259 (1989). 52. Audi, Religion and the Ethics of Political Participation (Review Essay), 100 ETmcs 386 (1990). 53. Richards, Book Review, 23 GA. L. REv (1989). 54. Nagel, supra note 50, at 216.

20 1990] SOME FURTHER THOUGHTS 1037 not sufficient to satisfy this higher-order impartiality. The inquisitor who orders that a heretic be burned may be treating the heretic impartially, given the inquisitor's belief about what is true. The inquisitor is interested in helping to save everyone's soul. Imposing that penalty may seem the way to save the most souls and perhaps even the heretic's. 55 Higher-order impartiality, however, involves beliefs that "can be shown to be justifiable from a more impersonal standpoint.''36 Thus, one must appeal to a common ground. "This means that it must be possible to present to others the basis of your own beliefs, so that once you have done so, they have what you have, and can arrive at a judgment on the same basis. ' ' 7 For Nagel, the element of coercion in the political order necessitates reference to this higher-order impartiality. To justify making people do things against their will, an especially stringent requirement of objectivity in justification must be imposed; 5 " people need to be "impartial not only in the allocation of benefits or harms but in their identification." 5 9 This requirement is not met when "part of the source of your conviction is personal faith or revelation-because to report your faith or revelation to someone else is not to give him what you have, as you do when you show him your evidence or give him your arguments.'"' 6 Nagel supposes that the inquisitor would be unable to justify coercion to the heretic without relying on personal faith or revelation. Further, Nagel thinks that in the present state of moral debate, certain moral conclusions, about issues such as abortion, sexual conduct, and the killing of animals for food, rest on "personal moral convictions" that fail the test of higher-order impartiality, even when those conclusions are not tied to religious premises. 61 Nagel concludes that the state should not restrict individuals' liberty with respect to such issues if they are reducible to "confrontations between personal moral convictions.' '62 Nagel recognizes coercion cannot await unanimous agreement to be justified. Disagreements exist, and persist, even when reasons are offered that lie in the public domain. People have been exposed to different testimony and arguments, they have had different experiences, and they assess evidence and arguments differently, so reasonable disagreements result. 63 But "the distinction between a disagreement in the common, public domain and a 55. In a sense, the heretic would also be treated impartially if he receives appropriate retributive punishment for heinous wrongful acts or thoughts. 56. Id. at Id. at 232 (emphasis in original). 58. Id. at Id. at Id. at 232. Nagel makes clear that his argument does not depend on skepticism about the possible truth of conclusions arrived at by means other than common grounds of reason; it depends only on the inappropriateness of coercing others on the basis of reasons other than publicly available arguments. Id. at Id. at Id. 63. Id. at 234.

21 1038 DEPA UL LA W REVIEW [Vol. 39:1019 clash between irreconcilable subjective convictions [is not] too rarefied to be of political significance. Judgment is not the same as faith, or pure moral intuition.' ' The central criticism of my position by David Richards is that I have failed to "draw the proper inference from the distinction between reason in ethics and reason in politics, namely, that reasonable ethical choices may very well not satisfy the requirement of reason in politics. ' "6 Values internal to religious traditions are not necessarily "shared in the wider society, and are not therefore publicly accessible to all in the way that Lockean political legitimacy requires for their enforcement on the community at large."6 ' The idea that political judgments should be based on shared values and publicly accessible ways of determining truth plays a significant part in John Rawls's political philosophy. When I did my original research, I concluded that Rawls probably thought that in a just society all political issues, or at least all issues of justice, should be resolved in this manner. 6 7 In his most recent work, Rawls develops the idea that when issues involve constitutional essentials, including, for his theory, matters relating to fair opportunity and distribution of wealth as well as basic liberties, political power should be exercised "only in ways that all citizens can reasonably be expected to endorse publicly in the light of their own common, human reason. ' ' 68 For other subjects of legislation that do not involve constitutional essentials, citizens need not limit themselves to such stringent requirements regarding relevant considerations (although it is not clear, even then, that reasons based on explicit religious premises are appropriate). When it is doubtful whether or not a highly divisive subject involves a constitutional essential-rawls offers abortion as an example-citizens should try to articulate their claims by reference to publicly accessible values, if that is possible. 6 9 Relying more heavily on specific notions of religious liberty and nonestablishment, Robert Audi reaches conclusions about the place of religious convictions in public debate similar to those of Nagel, Richards, and Rawls. He states that: [I]n a free and democratic society, people who want to preserve religious and other liberties should not argue for or advocate laws or policies that restrict human conduct unless they offer (or at least have) adequate secular (nonreligious) reasons to support the law or policy in question (where an adequate reason for a law or policy is a proposition whose truth is sufficient to justify it)... A secular reason is, roughly, one whose normative force 64. Id. at Richards, supra note 53, at Id. at See K. GREENAWALT, supra note 2, at Rawls, The Domain of the Political and Overlapping Consensus, 64 N.Y.U. L. REv. 233, 244 (1989). 69. Rawls took this position in discussion and in a work in draft at a seminar on November 9, 1989 of the Program for the Study of Law, Philosophy & Social Theory, New York University School of Law.

22 1990] SOME FURTHER THOUGHTS does not (evidentially) depend on the existence of God... or on theological considerations. 70 Audi is not certain whether his "principle of secular rationale ' 7 ' applies to all laws and policies, but it does apply to laws that restrict conduct. 72 Audi continues that people should not construct secular rationalizations when they are really persuaded by religious considerations, and they should not present to others as sufficient reasons they find less than persuasive. 73 He argues for a "principle of secular motivation, ' 74 which provides that "one should not advocate or promote any legal or public policy restrictions on human conduct unless one not only has and is willing to offer, but is also motivated by, adequate secular reason." '7 This is a stronger requirement than that imposed by Audi's principle of secular rationale. Suppose a Roman Catholic has some grasp of the naturalist arguments for giving the fetus full moral status upon conception. Although he believes on the basis of church authority that these arguments are persuasive, he is not certain whether he would find them persuasive if he considered them without regard to church authority. If he presents the naturalist arguments as a basis for restricting abortions, he has offered what he regards as adequate secular reasons, thereby satisfying Audi's principle of secular rationale. But he is not now motivated by the force of those naturalist reasons taken by themselves; he is not sure what he would conclude about abortion were he to disregard church authority. Thus, he fails the test of secular motivation. In order to comply with the principle of secular motivation, he must make an effort to determine whether, apart from the church's authority, he thinks the naturalist arguments are persuasive. 76 The "principle of secular resolution, ' 77 Audi's third principle, is related to his other two principles and requires that the final resolution of political issues be made along secular lines; decisions should be fully warranted by secular considerations. 78 These three principles apply to private citizens as well as public officials. 79 In Audi's view, they do not preclude some argument for political positions in terms of religious reasons, "though this has its dangers." 80 Richards also seems to approve of some public advocacy of moral views based on religious premises so that in "public testimony and debate" it can be seen "whether such views have an independent basis." 8 ' 70. Audi, supra note 51, at Id. at 279 (emphasis omitted). 72. Id. 73. Id. at Id. at Id. (emphasis in original). 76. See id. at Id. at Id. 79. Id. at Id. at Richards, supra note 53, at 18.

23 1040 DEPA UL LA W RE VIEW [Vol. 39:1019 Richards and Audi, like Nagel, believe that if the secular reasons, by themselves, are inconclusive about whether behavior should be restricted, then the behavior should be left free. Generously granting that my position is a plausible alternative to his, Audi challenges some of my central supporting arguments. He suggests that for some issues, the apparent radical inconclusiveness of secular considerations may derive from 2 the influence of religious considerations. He is hopeful that in the months or years appropriate for major public policy debates, rational secular resolutions for issues like abortion may be achieved. 83 This 4 is a theme echoed by David Richards, who also notes "a history of progress in moral theory." 85 Audi suggests that when a position is taken on religious grounds, a person is less likely to compromise and more likely to disapprove of opponents. 6 Moreover, if religious influences are perceived to be important determinants of public policy, some people will be offended or alienated 7 and social harmony will be undermined. Audi disagrees with my suggestion that asking religious believers to make decisions about public policy on secular grounds asks something that exceeds reasonable expectations about their capacities. 8 And so long as religious beliefs and their expression are tolerated, he sees no failure of toleration and fairness 9 in holding out the model of secular decision. B. Response To Proposals That People Should Try To Exclude Religious Convictions From Political Choices Any full response to the idea that officials and citizens acting politically should restrict themselves to shared values and publicly accessible reasons must depend on the more complete treatment in my book. Still, a sketch of my difficulties with some of the major points may suggest the main bases of disagreement. I think Nagel, Rawls, Audi, and Richards are correct in that membership in a liberal democracy involves a commitment to publicly accessible reasons, but I think the relevant commitment is different from and more limited than they suggest. Citizens and officials should remain open to publicly accessible reasons. They should generally make arguments in terms of these reasons, and they should resolve problems on the basis of those reasons when possible. But, with some frequency, even as to constitutional essentials and their applications, 9 0 people will be thrown back to the sources Nagel classifies as 82. Audi, supra note 52, at Id. 84. Richards, supra note 53, at Id. at Audi, supra note 52, at Id. 88. Id. at Id. at In my book, I develop this point with regard to levels of welfare. See K. GREENAWALT, supra note 2, at

24 1990] SOME FURTHER THOUGHTS 1041 personal moral convictions. Indeed, and this is critical, the manner in which people understand and employ shared reason is intertwined with the content of their personal convictions, including their religious convictions. 91 One problem is that the religious believer will find it difficult to put aside religious convictions when it comes to evaluating secular or publicly accessible reasons. This is illustrated by the devout Roman Catholic who tries to evaluate the naturalist arguments about abortion on their own terms. If one's confidence in religious authority leads one to the psychological certainty that a philosophical argument is convincing, it takes substantial mental agility to assess the argument separately from one's belief in the religious authority. While this sort of agility is highly prized in law schools, and perhaps philosophy departments, it is nevertheless an uncommon virtue. A yet more difficult problem concerns the place that religious premises play in forming the groundwork of one's thought about publicly accessible reasons. Suppose that Janet's Christian beliefs include ideas of ineradicable selfishness and the primacy of a "love ethic." She will bring those presuppositions to bear in evaluating many of the ordinary arguments about important laws and social policies, and they will color her discourse in the terminology that is common to all members of society. The love ethic is not separate from generally shared notions of equality and concern; instead, it gives a certain tincture to those notions. To ask Janet to try to erase these religious aspects of her understanding of secular arguments would be to ask that she attempt a division that is almost inconceivable. Moreover, if she were asked to start her reasoning process without her religious assumptions, she would be hard put to discover some other basic premises. I am not sure that each of the authors I have discussed actually thinks that Janet should make this effort to extract the religious underpinnings from her arguments, but that seems to be the implication of at least Audi's principle of secular motivation. It is especially for this mingling of religious premises and secular reasons that asking Janet to figure out if she has adequate secular reasons seems impossibly demanding. That is part of why I think liberal citizenship does not require people to make that attempt. Because this problem infects the essentials of constitutionalism and their applications, as well as ordinary political issues and issues that lie on the borders of the essentials of constitutionalism, the difficulty cannot be met by a division between constitutional and other political issues, with resolution of the first category limited solely to publicly accessible reasons. I want to address a few other critical difficulties with the recommendations that people restrict themselves to publicly accessible or nonreligious reasons for political decisions. Nagel's suggestion that grounds of coercion must be based on reasons that can be shown to be justifiable from an impersonal 91. By classifying religious convictions as personal in this sense, I do not mean that they are developed individually or in isolation from a religious community and tradition. All that matters for this purpose is that the convictions are not shared by members of society at large.

25 1042 DEPA UL LA W REVIEW [Vol. 39:1019 standpoint presents a serious problem of application. Suppose that religious believers are persuaded by Nagel's argument of political philosophy. How then are they to act? I have mentioned that no sharp distinction exists between faith and reasoned support for belief; what one is inclined to believe because of personal inspiration or participation in a religious tradition is subject to some tests of ordinary reason. Nagel can handle this complexity by saying that any belief that rests substantially on something beyond publicly accessible reasons is to be excluded. But what of those believers, mentioned in my last section, who think that all their important beliefs can be established by accessible reasons? Let me oversimplify crudely to suggest how biblical authority might be conceived as being grounded in this way. If there were overwhelming evidence that someone, according to every criteria of death, had died, had remained dead for some days, and had actually lived again in his or her earthly body, and if there were no plausible scientific hypothesis to explain such an event, "(common sense" would suggest that some supernatural force was at work. Of course, the presence of such a force would not be clear beyond doubt, because a natural scientific explanation might be forthcoming in the future. Nevertheless, the resurrection would constitute some evidence of a supernatural power. Although faith is the main basis for many Christians' belief in the resurrections of Jesus and Lazarus, many Christians believe the historical evidence found in the gospel accounts of Jesus's resurrection is a substantial reason for accepting Christianity. Some believe further that the general historical accuracy of the Bible, the miracles reported in it, and the fulfillment of biblical prophecies strongly indicate its powerful authority. These latter claims of reason are conceived to be publicly accessible, and I think they largely are. Nagel undoubtedly believes, and I do as well, that there is a limit to how far reason goes towards establishing claims about religious truth, but some Christians think common reason strongly supports fairly detailed Christian beliefs and their ethical and political implications. What are the appropriate limitations on the political conclusions these Christians rely upon and argue? In one sense they should not reason in terms of Christian belief because, assuming Nagel is right, such belief includes central personal convictions. But for Nagel to persuade these Christians to accept this limitation, he must show not only the soundness of his view about higher-order impartiality but also that they have misconceived the true grounding of their religious beliefs. Until they are persuaded on the latter point, their view of how Nagel's position on impartiality applies to them will not exclude all or many of their religious convictions. 92 The natural consequence will be that religious claims will be relied upon and argued for. The political debate will include arguments about the boundaries of common 92. Remember that Dean Gamwell's argument for allowing religious convictions in political dialogue is precisely that many religious convictions are subject to rational scrutiny. See supra notes and accompanying text.

26 1990] SOME FURTHER THOUGHTS 1043 reason and whether religious beliefs can be fully supported by that reason. While the flaws of claims that religious beliefs can be supported in this way may seem common currency among philosophers, and perhaps among most leading academics writing about religion, many believing Christians distrust both these groups. We should not suppose, then, that even universal acceptance of Nagel's idea of higher-order impartiality would itself quickly lead to the exclusion of religious claims from the political arena. An alternative approach would be simply to specify that all religious convictions fall into the excluded category. It might be argued that in our society there is a general understanding that religious beliefs are not sufficiently accessible to common reason or that principles of religious liberty and nonestablishment call for them being considered in this way. Robert Audi makes the second kind of argument. Under this approach, the disfavoring of religious reasons in comparison with secular ones need not rest on application of a general principle of higher-order impartiality; restraint about religious reasons is called for on the basis of historical tradition and a sound division between religion and politics in a liberal society. 9 " My assertion that a principle excluding religious convictions as a basis for political choice would be intolerant and unfair is primarily aimed at the view that these convictions warrant special exclusion, that they should be excluded to a degree even if other personal moral convictions are not. I do grant there are some special dangers when some religious premises figure in political choices, and I also grant that general public acceptance of my position might lead to overuse of religious premises according to my own standards. 94 But neither of these worries would warrant Audi's stringent principles of secular rationale, secular decision and secular motivation, if these principles were understood to permit a self-conscious input from personal moral convictions other than religious convictions. The unfairness would lie in effectively privileging personal moral convictions with a nonreligious source. 9 " We are after all a society with a deep religious tradition. Saying that citizens engaged in political choices may reach beyond publicly accessible reasons to all but religious sources would be odd. Further, such a principle would leave the religious person with the nearly impossible job of filling the gap. I have already suggested how hard it will be for believers to disentangle religious sources from deep premises of fact and value that underlie their use of 93. Let me remind the reader that the exclusion of religious convictions is by no means an obvious derivative of the principles of religious liberty and separation. Dean Gamwell and I are strong separationists about state involvement with religion and state sponsorship of religious beliefs; but each of us thinks that citizens and officials can bring religious premises to bear on their political choices, and Gamwell favors extensive public discussion on those terms. 94. On this point, see Schauer, May Officials Think Religiously, 27 WM. & MARY L. REv. 1075, (1986). 95. Robert Audi has suggested in personal conversation that he thinks at least many personal moral convictions of the sort I mean here would fail a test of adequacy for adequate secular reasons. The result would be that the possible unfairness I mention in the text would be eliminated or largely reduced in Audi's own full position about adequate reasons.

27 1044 DEPA UL LA W REVIEW [Vol. 39:1019 commonly accessible reasons. Imagining the personal moral convictions they would have were they without religious belief would be even harder. How does a devout Roman Catholic begin to guess how he would feel about the fetus were he not a religious person? Audi rejects my suggestion that principles of exclusion would be intolerant of religious belief. He points out that his approach is compatible with simple toleration of religious beliefs and their expression. People may be entirely free to hold beliefs and defend them, even if they are told to exclude them from political choice.96 What I mean by toleration, however, is more subtle. Full toleration, in my view, requires a grasp of the role that religious premises play in the lives of serious believers. Calling upon believers to perform impossible mental gymnastics and to give less effect to many of their deepest convictions than may people with other kinds of personal convictions represents a failure of consideration and toleration in this more subtle sense. 9 7 Nagel, Audi, and Richards all argue that when publicly accessible arguments are inconclusive, a " liberty should not be restricted. I believe there is a plausible argument that fundamental liberties, themselves supported by public reasons, should not be restricted without arguments that can also be supported "all the way down" by secular, publicly accessible reasons. I address variations of that position in my book and reject them with some hesitation." Here I want only to address the simpler, more general position about libertyrestricting laws and policies, the position that no restriction on liberty is warranted without full support in public reason. Nagel seems to suggest two related reasons for more stringent justification when liberty is restricted than when other laws and policies are involved. Coercion of individuals requires heavy justification and noncoercion is always an option. For other matters, including nuclear deterrence, direct coercion is not involved and policy must be determined one way or another.1 As far as coercion is concerned, Nagel's example illustrates a more pervasive difficulty. Soldiers must fight with the weapons assigned, and when there is a draft civilians are coerced to be soldiers. The indirect effect of a misguided nuclear policy could be death for coerced soldiers as well as fortuitously involved civilians. That would seem to demand more stringent justification than that required to prohibit hunters from shooting members of an endangered species, which represents a direct restriction on individual liberty. 96. Audi, supra note 51, at ; Audi, supra note 52, at The idea here matters much more than the word "toleration." If one thinks that toleration should be limited to the more simple sense that Audi uses, one may see my claim as one about unfairness and failure of consideration rather than intolerance. 98. Nagel and Richards plainly include cases when judgment must be based on nonreligious personal moral convictions. Nagel, supra note 50, at ; Richards, supra note 53, at From his essays on this topic, I am not sure whether Audi's requirement of adequate secular reasons is meant to exclude a broader category than religious convictions. See Audi, supra note 51, at 293. But see supra note See K. GREENAWALT, supra note 2, at See Nagel, supra note 50, at 233.

28 19901 SOME FURTHER THOUGHTS 1045 Moreover, almost any policy that requires money to implement it is paid for by taxes; coercing me to pay taxes for a program that cannot be justified on grounds accessible to my reason seems as much of an imposition as forbidding me to shoot members of an endangered species on grounds not accessible to my reason. If coercion were the key to a more stringent principle, one would have to look beyond the conceptual categorization of the particular law or policy to consider the actual effect on people. For many matters the government must choose one way or another, and it is true that any choice about nuclear deterrence risks harm. But it would be misleading to suggest that when common grounds are inconclusive, the state can comfortably leave people free of restrictions on individual behavior. Suppose common grounds are inconclusive about whether a species, of no apparent value to human beings, should be protected. Suppose, further, that the great majority of people think the species should be protected. Those who want to hunt the species for fun contend that their liberty should not be restricted except on grounds that appeal to their reason. A choice in favor of human liberty is a choice to allow the species to be exterminated. For more typical issues of "animal rights," the choice in favor of liberty is a choice to leave defenseless creatures without protection. In the modern welfare version of political liberalism, no standard principle in favor of liberty of the stronger is warranted when neither of the competing claims is fully supported on commonly accessible grounds. IV. CONCLUSION The positions that I have considered in this lecture share three important features. First, citizens and officials making political choices are to act according to roughly the same principles of appropriate reason and argument. 01 Second, the appropriate grounds to rely upon are regarded as reasonably accessible in some sense. 0 2 Third, there should be a substantial correlation between the actual reasons for choice and the arguments put forward on behalf of choice. 03 My position conforms with the others in closely aligning appropriate grounds for officials and citizens. But I am somewhat chastened to find myself at odds with positions on both ends of the spectrum on the other two features. I admit more play for personal grounds of choice than any of the other positions. Recall that Gamwell, Perry, and Lovin, who defend reliance on religious convictions and argu But see Schauer, supra note 94, at Schauer contends that although citizens in certain instances may properly rely on religious convictions in taking political positions, it would be inappropriate for public officials to rely on religious grounds in deciding political issues. Id Lovin and Perry apparently would permit religious arguments that are not commonly accessible, but they have confidence about the discussion of religious views as aspects of political dialogue Audi, and apparently Richards, would permit some discussion of religious arguments when a person's choice can be fully justified on other grounds.

29 1046 DEPA UL LA W REVIEW [Vol. 39:1019 ments based on them, think that these convictions are more broadly accessible to nonbelievers than I have assumed. Although they have a different view of religious convictions from Richards and Audi, they share with them a confidence about what common reason can resolve that is considerably greater than mine. The degree of noncorrelation in my position between reliance on religious convictions and argument in those terms has always disturbed me. My defense is that the divisive potential of reliance on religious conviction is much softened by political argument in secular terms. Further, asking people to present positions in secular terms does not require the extreme mental agility of asking that they try to decide without reliance on religious premises that color their views. A precept of secular argument is much more achievable than a precept of reliance on secular reasons. The contention that people should argue on the basis of the considerations that actually underlie their political positions is undoubtedly forceful. I find it compelling within small groups. However, I believe argument for political positions in the wider polity is more like legal argument than a candid account of all that one finds relevant. One makes arguments likely to persuade and reassure the audience, and one makes arguments in terms that affirm more general political values. I do think people should acknowledge the place of religious convictions in their own positions, but I still believe it is counterproductive for debates on particular political issues to be formulated explicitly in terms of religious convictions. A broader examination of religious positions and their implications in the general culture would be desirable. Dialogue about religious questions has an important place, but I still do not think that place is in debates focused on the wisdom of particular laws and policies.